Bug#474065: Acknowledgement (Lift the 64000-byte restriction on the size of messages sent to subprocesses.)
I mentioned a bug in ReadMessages in the initial submission. I'm not
sure whether or not this is really-and-truly a bug, though. It looks to
me like the code that uses ReadMessages expects to just call it over and
over when input is ready, so as long as we don't actually drop part of a
message (what my patch fixes), I would expect that it's OK if
ReadMessages occasionally exits before consuming all the available
input.
Daniel
Reply to: